Global Fit Inference for LISA with samsara++
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Astrophysics. The day in astrophysics: observations, models and what the data forces.
Vera: I'm Vera.
Jocelyn: And I'm Jocelyn, here with Subrahmanyan.
Vera: We're looking at the paper: "Global Fit Inference for LISA with samsara++".
Jocelyn: Here's the gist of the paper. The paper presents samsara++, a C++/CUDA sampler designed to address the trans-dimensional Global Fit (GF) problem for LISA by combining birth-death processes with mutation moves within a continuous-time Markov chain framework This algorithm is crucial because it allows for the joint inference of an unknown number of sources, parameters, and instrumental noise in LISA data, which overlap in both time and frequency.
Vera: The core methodology The approach utilizes a birth-death process (BDP) within a continuous-time (CT) MCMC framework to explore the trans-dimensional parameter space by proposing new states with a larger (birth), smaller (death), or unchanged (mutation) number of sources.
Jocelyn: The transition rates are chosen to satisfy detailed balance, analogous to the Metropolis–Hastings acceptance ratio.
Vera: The sampler is designed to address the challenge of combining high dimensionality with uncertainty in the number of sources, which requires methods capable of exploring models of different dimensions.
Title and authors: Vera: So, to wrap up what we just talked about, this paper introduces samsara++, which is this new C++ CUDA sampler for LISA data that tackles that whole global fit problem—getting the number of sources, their parameters, and the noise all at once.
Jocelyn: Yeah, basically it uses a continuous-time Markov chain with birth-death processes to propose different numbers of sources, like adding one more or removing one less. It's a way to explore that huge parameter space by following a path that’s driven by the posterior distribution itself.
Subrahmanyan: From my side, what I find interesting is how they handle the high dimensionality; they use specific optimizations, like indexed storage in C++ and CUDA to save memory when things are changing. That’s a big engineering win for running this kind of thing on real hardware.
Vera: And not just that, they’re using parallel tempering to help the sampler explore different regions of the result space better. It’s designed to help the chain escape getting stuck in local modes where there might be too many redundant sources.
Jocelyn: Right, and they tested it on mock data and showed it can recover the number of sources injected without needing any prior knowledge about how many there should be. That’s a strong result for figuring out if we’re missing something in our source counts.
Subrahmanyan: It means this method could be the foundation for doing a proper global fit, which is what LISA really needs to do—jointly modeling everything from Galactic binaries to extreme mass ratio inspirals.
Vera: So, they're moving away from the old methods like RJ-MCMC by following a trajectory through all those dimensions simultaneously. It’s about making the inference process itself smarter and more efficient than just running separate fits for each thing.
Jocelyn: And I think what this shows is that we can tackle problems where the model structure itself, like the number of sources, is unknown, instead of having to guess it ahead of time.
Subrahmanyan: Exactly. This opens up possibilities for combining different types of signals—like foregrounds and astrophysical sources—into one coherent picture instead of treating them separately.
Vera: It’s a powerful tool for testing how well we can actually fit the data when the complexity is this high, which is exactly what LISA will be dealing with.
Jocelyn: We need to see if this efficiency holds up when we start feeding it all that real noise and all those different source populations into the model.
The paper's summary: Vera: So, to recap, samsara++ is this new sampler for LISA data that uses birth-death processes to handle the problem of figuring out unknown source numbers and noise all together in a global fit analysis.
Jocelyn: Yeah, it’s using that continuous-time Markov chain idea to propose different numbers of sources at each step, like birth or death moves. It’s designed to explore those complicated parameter spaces without having to guess the number of sources first.
Subrahmanyan: From a theoretical side, what I like is how they model the signals—they use mildly chirping sinusoids for Galactic binaries and account for Doppler modulation in the waveform parameters like frequency and amplitude. That’s a physically motivated way to build the signal model.
Vera: And they’re using heterodyning to speed up those time-domain calculations, which cuts down on the number of operations needed to calculate likelihoods by a factor related to the frequency points. That makes it computationally feasible for real data sets.
Jocelyn: Right, and they’ve also made some big computational improvements, like using C++ CUDA for parallel computation and an indexed storage scheme so memory doesn't get wasted when the population sizes change. That’s crucial for handling high-dimensional problems.
Subrahmanyan: I’m interested in their proposal to use a trans-dimensional parallel tempering scheme to improve mixing. If the sampler can explore different temperature chains well, it means we have a better chance of finding the true posterior mode without getting stuck in weird local configurations.
Vera: So they aren't just presenting one method, they are suggesting ways to make this sampler even more robust and efficient by adding these specific computational tricks and sampling techniques.
Jocelyn: And there’s the idea that they could extend this further by using parallel evolution for different species of sources. If you have many non-overlapping frequency bands, you can run those populations in parallel, which is a great speedup.
Subrahmanyan: That sounds promising for scaling up the model; if we can evolve those different source groups simultaneously, it gets closer to modeling the full complexity of LISA’s expected input.
Vera: So what this suggests is that we aren't just looking at one fix, but a whole pipeline of improvements—from the core sampler design to how they handle the signal modeling and how they run the code efficiently.
Jocelyn: And it points toward a future where we can apply this framework not just to DWDs, but to include all those other sources we’ve been discussing, like MBHBHs and EMRIs, in that single global fit.
The paper's improvements: Vera: So we've covered how samsara++ uses birth-death processes to do that trans-dimensional global fit for LISA data, figuring out source counts, parameters, and noise all at once.
Jocelyn: Yeah, it shows a way to explore those huge parameter spaces by proposing moves that add or subtract sources from the current model state. It's about following the posterior distribution without knowing exactly how many sources there are beforehand.
Subrahmanyan: From my perspective, the modeling of those signals as mildly chirping sinusoids is pretty standard, but I appreciate how they’ve tied their spectral width to things like observation time and Doppler modulation. That keeps the physics grounded in what we actually see.
Vera: And that whole process is made feasible because of the computational efficiency gains they made, using parallelization and those clever indexing schemes to keep memory management clean as the chain runs.
Jocelyn: I think it really opens up possibilities for combining all these different signal types—the Galactic binaries, the foregrounds, and the exotic sources like MBHBHs—into one coherent analysis. It moves us closer to a full global fit model.
Subrahmanyan: If this method can scale up to handle all those source species without losing its efficiency, it becomes a real tool for testing cosmological models where we need that level of joint inference.
Vera: It’s a solid piece of work because they've shown that the sampler can recover the number of injections even when starting from an empty state, which is exactly what we need to test our analysis pipelines.
Jocelyn: And it gives us a clear path forward for LISA data analysis, showing how to handle this kind of complexity in a practical way.
Subrahmanyan: Ultimately, the success of this Global Fit Inference for LISA with samsara++ really depends on whether we can successfully extend it to include all those different source populations. That’s the next big challenge for structure formation studies.
Conclusion: Vera: So we're wrapping up on "Global Fit Inference for LISA with samsara++," which is this new C++ CUDA sampler that tackles figuring out source counts, parameters, and noise all at once for LISA data.
Jocelyn: Yeah, it shows a way to explore those huge parameter spaces by proposing moves that add or subtract sources from the current model state. It's about following the posterior distribution without knowing exactly how many sources there are beforehand.
Subrahmanyan: From my side, I think what they did well is how they handle the high dimensionality; using indexed storage in C++ and CUDA to save memory when population sizes change is a big engineering win for running this kind of thing on real hardware.
Vera: And that parallel tempering setup they introduced really helps them sample better by letting the algorithm explore more probable configurations without getting stuck in weird local spots.
Jocelyn: Right, and the results with those mock analyses showed that the sampler can recover injected source numbers even when starting from an empty state, which is a strong result for figuring out if we're missing something in our source counts.
Subrahmanyan: It means this method could be the foundation for doing a proper global fit, which is what LISA really needs to do—jointly modeling everything from Galactic binaries to extreme mass ratio inspirals.
Vera: They’re moving away from those older RJ-MCMC methods by following a trajectory through all those dimensions at once, which makes the inference process itself much more efficient.
Jocelyn: And I think what this shows is that we can tackle problems where the model structure, like the number of sources, is unknown, instead of having to guess it ahead of time.
Subrahmanyan: Exactly. This opens up possibilities for combining different signal types—like foregrounds and astrophysical sources—into one coherent picture instead of treating them separately.
Vera: It’s a solid piece of work because they've shown that the sampler can recover the number of injections even when starting from an empty state, which is exactly what we need to test our analysis pipelines.
Jocelyn: And it gives us a clear path forward for LISA data analysis, showing how to handle this kind of complexity in a practical way.
Subrahmanyan: Ultimately, the success of "Global Fit Inference for LISA with samsara++" really depends on whether we can successfully extend it to include all those different source populations. That’s the next big challenge for structure formation studies.
Dipartimento di Fisica “E. Fermi”, University of Pisa · INFN, Sezione di Pisa · Gran Sasso Science Institute (GSSI) · INFN, Laboratori Nazionali del Gran Sasso · Dipartimento di Fisica “G. Occhialini”, University of Milano-Bicocca · Institute for Gravitational Wave Astronomy & School of Physics and Astronomy, University of Birmingham
astro-ph.IM, astro-ph.CO, gr-qc
Submitted: 2026-11-12
Updated: 2026-11-12
Comments: 23 pages, 9 figures, 2 tables, 3 appendices
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 92/100
The gist: The paper presents samsara++, a C++/CUDA sampler designed to address the trans-dimensional Global Fit (GF) problem for LISA by combining birth-death processes with mutation moves within a…
Key concepts
- Global Fit (GF)
- A technique needed for LISA data where signals from many different gravitational wave sources overlap in both time and frequency. GF aims to jointly model these overlapping signals, including resolvable sources, stochastic foregrounds, and instrumental noise.
- Birth-Death Process (BDP)
- This is a mathematical process used in samsara++ to propose new states in the parameter space. It models the evolution of the number of sources by proposing 'birth' (adding a source), 'death' (removing a source), or 'mutation' (changing parameters) moves.
- Parallel Tempering (PT)
- A method used within samsara++ to improve exploration of complex parameter spaces. It runs multiple MCMC chains at different temperatures simultaneously, allowing the algorithm to jump out of local optima and explore more probable configurations effectively.
Terminology
Summary
The paper presents samsara++, a C++/CUDA sampler designed to address the trans-dimensional Global Fit (GF) problem for LISA by combining birth-death processes with mutation moves within a continuous-time Markov chain framework This algorithm is crucial because it allows for the joint inference of an unknown number of sources, parameters, and instrumental noise in LISA data, which overlap in both time and frequency. p. 0
The core methodology
The approach utilizes a birth-death process (BDP) within a continuous-time (CT) MCMC framework to explore the trans-dimensional parameter space by proposing new states with a larger (birth), smaller (death), or unchanged (mutation) number of sources. The transition rates are chosen to satisfy detailed balance, analogous to the Metropolis–Hastings acceptance ratio. p. 1 The sampler is designed to address the challenge of combining high dimensionality with uncertainty in the number of sources, which requires methods capable of exploring models of different dimensions. p. 2
Computational optimizations
To maximize computational efficiency, samsara++ employs several optimization strategies. These include shared-memory parallelism using C++/CUDA to compute rates in parallel and a custom indexing scheme to avoid redundant memory allocations for arrays whose size could change. p. 1 Furthermore, the algorithm exploits parallelism by simultaneously evolving DWD populations in independent frequency bands and by introducing a trans-dimensional implementation of parallel tempering (PT). p. 1 Amortized data structures are used to avoid reallocating memory every time the dimension of the parameter space changes, reducing the memory allocation cost from O(Ngen) to O log2 N max pop. p. 7
Inference modeling
The likelihood function for DWD signals is modeled in the Fourier domain, assuming a zero-mean, stationary auto-covariance process for instrumental noise. The authors adopt an approximation scheme for computing rates by considering the variation of the log-likelihood due to adding or removing a single source. p. 3 Waveforms for DWDs are modeled as mildly chirping sinusoids, with their spectral width determined by accounting for the finite observation time and orbital Doppler modulation. p. 4
Validation and results
The sampler's performance is validated using three mock GF analyses of DWD signals and instrumental noise. The tests demonstrate that the sampler can recover the injected number of sources starting from initial empty states, without using information from the injections or astrophysical expectations to inform the prior on the number of sources. p. 1 In a stress test involving 866 binaries, samsara++ recovers different numbers of injections in approximately the same number of generations as other runs. p. 0 The final test case with 131,480 source parameters shows the posterior mode of the source count matches the injected count in approximately 83% of the 250 bins. p. 0
Conclusion
samsara++ represents a paradigm shift from RJ-MCMC and product-space methods by following a posterior-driven trajectory in the trans-dimensional parameter space. The algorithm's efficiency stems from its custom data structures, parallelization across three nested levels, and the use of PT to improve mixing and convergence. The work provides a foundation for extending samsara++ toward a comprehensive LISA global fit by including Galactic binaries, foregrounds, MBHBs, and EMRIs in the source model. p. 0
Pruning redundant reasoning tokens
The Laser Interferometer Space Antenna (LISA) will observe a superposition of gravitational-wave signals, requiring joint inference of an unknown number of sources, parameters, and instrumental noise. To disentangle these contributions, that overlap in both time and frequency, a Global Fit (GF) analysis is needed. p. 0
How the training signal is built
The posterior distribution is given by Bayes’ theorem p(Θ d, H) = π(Θ H)L(d Θ, H) Z(d, H), where d are the data observed by LISA and H is the model hypothesis for the signals we are considering. The prior distribution π encodes the previous knowledge of the distribution of Θ for the model H, while the likelihood function L represents the probability density of the data given Θ and H. p. 2
Where the gains come from
The computation of birth, death, and mutation rates is an intrinsically parallel operation, allowing for speedup in computing time and number of steps by increasing the number of proposals at each iteration. The C++/CUDA version of samsara++ is specifically designed for the GF and uses an optimized indexing scheme to avoid redundant memory allocations for arrays whose size could change. p. 1
The LISA data streams will contain a superposition of GW signals from many sources of different classes, together with instrumental noise. In the samsara formalism developed in [58], we refer to a single source of GWs as an individual, while we identify as species the possible classes of signals. p. 1
The society specifies all the information about the state of the CT-MCMC at a given step. In a Bayesian framework, GF inference aims to determine the joint posterior distribution of the society parameters Θ. p. 2
The next state in the chain is selected among the candidate states from a multinomial distribution determined by the rates. If a birth or a death is selected, the new point in the parameter space is always accepted, while if a mutation process is selected, the new point is accepted or rejected with the MH scheme. p. 2
The LISA data streams will contain a superposition of GW signals from many sources of different classes, together with instrumental noise. Among the variety of deterministic sources which could be resolved by LISA, there are systems of massive black hole binaries (MBHBHs) [19], extreme mass ratio inspirals (EMRIs) [20], and compact binary systems of stellar mass in an earlier stage of evolution with respect to the ones detected by ground-based interferometers [21, 22]. p. 0
The overlap of these signals in both time and frequency poses a major data-analysis challenge for LISA, underscoring the need for a global fit (GF) that jointly models resolvable sources, stochastic foregrounds, and backgrounds.
The posterior evaluations and rate calculations in Eq. (3) can be parallelized across proposals, since each candidate is evaluated relative to the same current state. p. 2 The proposals adopted in samsara++ for the DWD species are illustrated in Figure 7. p. 2
In this work, we apply samsara to perform efficient trans-dimensional Bayesian inference over DWD populations and instrumental noise. These tests highlight the code performance on large datasets and high-dimensional parameter spaces. p. 3
We model the noise process as a zero-mean, stationary auto-covariance process, thus, according to the maximum entropy principle, the likelihood of the data is a multivariate Gaussian distribution. The quantity SI is defined in terms of the PSD of the noise as SI (f, Θn) ≡ PSD(f, Θn) 4∆f. p. 3
To overcome this issue, we adopt an approximation scheme for the computation of the rates of DWDs by considering the variation of the likelihood. The computational cost of samsara for DWD is O NDWDNchNhet and O NchNhet operations for updating the rates and log-likelihood respectively. p. 3
Galactic binaries emit GWs in the LISA band during the early stage of their inspiral and can be modeled as mildly chirping sinusoids. For this reason, following [33, 88–91], for DWD we consider the parameters θDWD ≡ n fGW, ϕ0, A0,sin β, λ, cosι, ψ, ˙fGWo. p. 4
The GW phase of the chirping sinusoids is given by ΦGW(θDWD, t) ≡ ϕ0 + 2πfGWt + π ˙fGWt2. The two GW polarizations are H˜+(θDWD, t) =A (1 + cos2(ι))2 cos [ΦGW(θDWD, t)], H˜×(θDWD, t) =A cosι sin [ΦGW(θDWD, t)]. p. 4
The computation of the waveform in the time domain requires evaluating Eq. (16) at the Nt points of the time series. p. 4 Heterodyning speeds up the computation of waveforms by a factor Nt/Nf t, while restricting the DWD signal to f accelerates the computation of the likelihood by a factor Nf /Nhet, where Nhet is the number of frequency points of the heterodyned DWD signal. p. 4
For the instrumental noise of LISA, we consider both the test-mass acceleration and the Optical Metrology System (OMS) noises. The PSD of these two contributions can be written in terms of two amplitudes as Spm(f) = 1 + 0.4 mHz f2
8 mHz4 and Sop(f) =
1 + 2 mHz f4" c2/2πc2.
Following [42], we adopt the equal-arms approximation for the three LISA channels, and identical acceleration and OMS noise PSDs across all links. The noise PSDs in the X, Y, and Z channels of LISA are identical; therefore, in the AET basis, the PSD of the instrumental noise is diagonal. p. 5
Since in our analysis we neglect data gaps, glitches, and sensitivity drift over time [11], the noise parameters involved in our Bayesian analysis are just Sacc and Soms. The noise parameters involved in the inference are θnoise ≡ racc, roms. p. 5
To characterize the Galactic DWD population, we specify a joint probability distribution for the source parameters introduced in Eq. (12), together with a probability mass function for the total number of systems. p. 5 These distributions are taken from [103, 104] and the catalog used to generate the injected signals is similar to that used in the LDC-2a “Sangria” dataset [105]. p. 5
The astrophysical catalog is obtained using the population synthesis code SeBa [25, 106–108], which evolves binaries from the Zero-Age Main Sequence (ZAMS) until the DWD formation and is calibrated against recent DWD observations. The full catalog contains approximately 2×10 7 sources, among which roughly 2 × 10 4 have SNR larger than 8 for one year of observations with LISA. p. 5
The prior on the number, here we use a uniform prior to be completely agnostic on the number of sources. The proposal distributions used in this analysis are illustrated in Figure 7. p. 2
In trans-dimensional inference, the prior π(ΘH), introduced in Eq. (1), includes a contribution that depends explicitly on the number of sources in each population. p. 5 It is reasonable to assume different species and individuals for each population to be independent; hence we can factorize the prior as π(ΘH) = N Ysp α=1 [π(Npop,αH) N Ypop,α i=1] [π(θi,αH)]. p. 6
The individual-source parameter priors are specified in Section III E. The prior we choose in our GF pipeline and the one adopted in all the test cases in Sections V, VI, VII. p. 6
We adopt a phenomenological source-count prior that penalizes additional sources, with a penalty motivated by their signal-to-noise ratio. The choice of ρthr can affect both the inferred source count and the efficiency of posterior sampling. p. 6
As discussed in Appendix C, when the parameter space is complicated, noise fluctuations and overlapping signals can induce spurious signals that cannot be associated with individual injected sources. These artifacts may change the source-count posterior or simply extend the burn-in phase. p. 6
The authors are grateful to A. Klein for the multiple helpful conversations and for providing the waveforms used in this work. p. 7 p. 7 p. 7 p. 7 p. 7 The authors thank S. p. 9 p. 9 p. 9 p. 9 p. 9 Babak, E. Barausse, M. Crisostomi, G. Demasi, C. J. Moore, C. Pitte, J. Pomper, R. Rosso, A Sesana and the participants of the conference “Listening to the Cosmos: New Frontiers in Gravitational Wave Physics” held at Galileo Galilei Institute for Theoretical Physics for useful discussions and comments. p. 0
The authors are grateful to A. Klein for the multiple helpful conversations and for providing the waveforms used in this work. The authors thank S. Babak, E. Barausse, M. Crisostomi, G. Demasi, C. J. Moore, C. Pitte, J. Pomper, R Rosso, A Sesana and the participants of the conference “Listening to the Cosmos: New Frontiers in Gravitational Wave Physics” held at Galileo Galilei Institute for Theoretical Physics for useful discussions and comments. p. 0
The authors are grateful to A. Klein for the multiple helpful conversations and for providing the waveforms used in this work. The authors thank S. Babak, E. Barausse, M Crisostomi, G Demasi, C J Moore, C Pitte, J Pomper R Rosso A Sesana and the participants of the conference “Listening to the Cosmos: New Frontiers in Gravitational Wave Physics” held at Galileo Galilei Institute for Theoretical Physics for useful discussions and comments. p. 0 p. 0 p. 0
The authors are grateful to A. Klein for the multiple helpful conversations and for providing the waveforms used in this work. The authors thank S. Babak, E. Barausse, M Crisostomi, G Demasi, C J Moore, C Pitte, J Pomper R Rosso A Sesana and the participants of the conference “Listening to the Cosmos: New Frontiers in Gravitational Wave Physics” held at Galileo Galilei Institute for Theoretical Physics for useful discussions and comments.
The authors are grateful to A. Klein for the multiple helpful conversations and for providing the waveforms used in this work. The authors thank S. Babak, E. Barausse, M Crisostomi, G Demasi, C J Moore, C Pitte, J Pomper R Rosso A Sesana and the participants of the conference “Listening to the Cosmos: New Frontiers in Gravitational Wave Physics” held at Galileo Galilei Institute for Theoretical Physics for useful discussions and comments.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Implement a Continuous-Time Markov Chain sampler based on birth-death processes to handle trans-dimensional Global Fit problems, as described in Section II and Section V. This allows the system to jointly infer an unknown number of sources, parameters, and instrumental noise by proposing new states with larger (birth), smaller (death), or unchanged (mutation) numbers of sources.
-
Enhance computational efficiency by utilizing C++ and CUDA for shared-memory parallelism, specifically exploiting
indexed storage
to avoid redundant memory allocations when birth and death take place. This reduces the complexity of removing a DWD from O(NchNhet) to O(1). -
Improve exploration of multimodal posteriors by integrating a trans-dimensional parallel tempering scheme, as detailed in Section IV D. This allows the algorithm to evolve an ensemble of chains at different temperatures, helping the cold chain escape configurations containing redundant sources and explore more probable configurations.
-
Implement amortized data structures using flattened one-dimensional C++ vectors with indexed dynamic arrays to avoid reallocating memory when the dimension of the parameter space changes. This reduces memory allocation from O(Ngen) to O log2 N max pop, where Ngen is the total number of CT-MCMC iterations and N max pop is the largest population size explored by the chain.
-
Enable parallel evolution steps for non-overlapping species using a Gibbs-like step adapted to CT-MCMC. By dividing the LISA frequency band into independent bins, this allows populations whose signals do not overlap in frequency to be evolved in parallel, which is described as a
great speed up
in exploration of the trans-dimensional parameter space. -
Develop an automated postprocessing pipeline using PETRA for catalog generation and signal identification. This pipeline should apply the Hungarian algorithm to find the best match of injections with singlesource posteriors in the catalog, improving code efficiency through alternative data structures and OpenMP parallelization.
-
Introduce a source-count prior that penalizes additional sources, motivated by their signal-to-noise ratio as described in Section III F. This prior is defined as π(Npop,αH) = 1 - e(-ρthr/2 exp(-Npop,α squared thr/2)), where ρthr > 0.
Sources
- Observation of Gravitational Waves from a Binary Black Hole Merger
- GWTC-1: A Gravitational-Wave Transient Catalog of Compact Binary Mergers Observed by LIGO and Virgo during the First and Second Observing Runs
- Population Properties of Compact Objects from the Second LIGO-Virgo Gravitational-Wave Transient Catalog
- GWTC-3: Compact Binary Coalescences Observed by LIGO and Virgo During the Second Part of the Third Observing Run
- GWTC-4.0: Updating the Gravitational-Wave Transient Catalog with Observations from the First Part of the Fourth LIGO-Virgo-KAGRA Observing Run
- The NANOGrav 15-year Data Set: Evidence for a Gravitational-Wave Background
- The second data release from the European Pulsar Timing Array III. Search for gravitational wave signals
- The Parkes Pulsar Timing Array Third Data Release
- Searching for the nano-Hertz stochastic gravitational wave background with the Chinese Pulsar Timing Array Data Release I
- Laser Interferometer Space Antenna
- LISA Definition Study Report
- Astrophysics with the Laser Interferometer Space Antenna
- Cosmology with the Laser Interferometer Space Antenna
- New Horizons for Fundamental Physics with LISA
- Report on the second Mock LISA Data Challenge
- The Mock LISA Data Challenges: from Challenge 3 to Challenge 4
- A LISA Data-Analysis Primer
- Science with the space-based interferometer eLISA. I: Supermassive black hole binaries
- Science with the space-based interferometer LISA. V: Extreme mass-ratio inspirals
- The promise of multi-band gravitational wave astronomy
Related papers
- On-sky demonstration of a vector Zernike wavefront sensor in a cascaded adaptive optics system
- Euclid Data Release 1 (DR1): Calibration strategy and observations
- The General Antiparticle Spectrometer (GAPS) Antarctic Balloon Payload
- Detection and excision of radio-frequency interference in radio arrays
- Unified matched-filter and excess-power pipeline for gravitational-wave detector characterization
- Ushering in sub-arcsecond angular resolution observatories in the far-infrared: The case for a double-Fourier interferometer